Stable Diffusion 3.5: Medium, Large and Turbo Models
ComfyUI Wiki
Stability AI's SD 3.5 open image models: Medium, Large, Turbo, and related ControlNets.
S
Stable Diffusion 3.5
Text-to-ImageOpen SourceMMDiTStability AI open image models with MMDiT architecture — strong text rendering and multi-subject composition. Use Stability AI checkpoints for FP16; Comfy-Org for FP8 all-in-one / scaled packs.
| Developer | Stability AI |
| Architecture | MMDiT — CLIP-L / CLIP-G / T5-XXL text encoders |
| License | Stability-AI-Open-RAIL-M |
| Native resolution | 1024×1024 (multi-aspect supported) |
| Variants | Medium (2.5B), Large (8B), Large Turbo (4-step), Comfy-Org FP8 |
| ComfyUI guide | /en/tutorial/advanced/stable-diffusion-3-5-comfyui-workflow |
Guides and workflows related to this model series.
SD3.5
Stable Diffusion 3.5 Workflow Tutorial in ComfyUI
Stable Diffusion 3.5 Workflow Tutorial in ComfyUI
Run Stable Diffusion 3.5 in ComfyUI: download SD3.5 Large, Large Turbo or Medium checkpoints and text encoders, then use the official FP16 or low-VRAM FP8 workflows.
SD3.5
Comments
Sign in with GitHub to join the discussion.